Epilepsy is a devastating illness affecting 3 million Americans. Unfortunately, we currently have only a rudimentary understanding of the intertwined issues of how to define the cortical areas which generate seizures and how those seizures start and spread. Based on preliminary data we posit that within the seizure onset zone epileptiform activity arises from deep cortical layers and then spreads via cortico-cortical connections to superficial layers. There is, consequently, a discernable physiological signature of the seizure focus and events leading to seizure initiation and propagation. We will test this hypothesis by recording synaptic activity, intrinsic currents, and action potential firing from all layers of human cortex during and between seizures using unique microelectrode arrays. Specifically, we aim to: 1. Demonstrate that the intracolumnar dynamics of interictal discharges depend on the location of that column in the epileptogenic network. We hypothesize that interictal discharges in the epileptogenic focus are generated by current sinks and increased neuronal firing in deep cortical layers, whereas propagated epileptiform discharges will show initial sinks and activation in middle and upper cortical layers. Such results are consistent with epileptiform activity arising from recurrent excitatory activity in deep cortical layers augmented by rebound intrinsic currents and delineate a microphysiological signature of ictogenic cortex. 2. Determine the role of different cortical layers and neuronal firing during seizure initiation. We expect that action potential firing in deep cortical layers within the seizure focus precedes overt seizure initiation. Further, we expect that these same layers are the site of current sinks during discharges that occur at seizure initiation. These features further define the seizure focus, shed light on how seizures start, and may provide a novel method for seizure prediction. 3. Examine the role of neuronal group dynamics during seizure spread. Finally, we hypothesize that from the focus, seizures spread by direct recruitment via projections to upper cortical layers. Further, for certain regions there will be increased involvement of deeper cortical layers as the seizure progresses correlated with an ability of that region to independently generate epileptiform discharges. Consistent with this evolution from direct recruitment to multi-focal autonomous event generation, analysis of functional coupling between cortical regions will show progression from tight to loose association. This description may further differentiate the seizure focus and suggest new strategies for interrupting seizure propagation. These aims address essential aspects of the neurophysiology of human seizures at an unprecedented level of detail and breadth. The results will lead to a clear mechanistic understanding of what constitutes the seizure focus in humans. This can lead to increased effectiveness of surgical management of medically refractory epilepsy, as well as innovative approaches to seizure prediction, detection and termination.